Introduction

Conventionally managed Sanqi depends on pesticides and synthetic fertilizers, resulting in potential consequences such as soil acidification1, nutrient depletion2, reduced soil microbial populations3, lowered soil enzyme functionality4, and soil-borne diseases5. These factors have negative effects on the quality and yield of Sanqi, further resulting in continuous cropping obstacles (CCOs). The organic management of Sanqi from the forest understory can increase pine-associated microbes6, improve soil quality6, enhance the quality of Sanqi7, and reduce the CCOs8, which has been widely promoted in Yunnan Province, China8. Although Sanqi from the forest understory has good quality, its yield is often lower than that of conventionally managed Sanqi. Prior studies have primarily focused on fertilization/foliar fertilization application2, planting density9, and CCOs prediction10 of the Sanqi from the forest understory. Therefore, enhancing the yield of Sanqi is an important issue that requires comprehensive investigation. Utilizing organic fertilizer in a strategic manner is currently identified as a significant approach to mitigate the challenges associated with Sanqi cultivation. This practice can effectively diminish soil-borne diseases, enhance the diversity of beneficial microorganisms, boost soil fertility, stimulate plant root growth, and enhance plant resilience to environmental stressors11.

Organic fertilizers comprise animal, plant, microbial, and compound biological organic fertilizers11. Humic acid, a recently developed fertilizer, plays a significant role in maintaining soil health and facilitating the absorption of nutrients by plants12,13. Research has demonstrated its ability to stimulate plant growth14, boost the availability of nutrients in soil15, increase the activity of soil enzymes4,16, and influence the diversity of soil microbes15,17. HA can increase the net photosynthetic and transpiration rates of Vicia faba14 and Vigna radiata18, thereby promoting plant growth. Simultaneously, nitrogen, phosphorus, potassium, and organic matter were significantly increased in the soil of some plants, such as Arachis hypogaea15, Musa nana Lour19, Gossypium hirsutum20, Triticum aestivum21 following the HA application. Furthermore, the addition of HA can lead to an increase in fungal abundance in Arachis hypogaea soils15, and in Firmicutes, Basidiomycetes, and Mortispora in Dioscorea esculenta soils, respectively22. HA increases the diversity of bacteria, fungi and actinomyces in Dioscorea esculenta23, Arachis hypogaea15 and Pisum sativum soils17. Furthermore, the application of HA not only alters the community structure and species composition of soil fungi but also significantly decreases the presence of soil-borne pathogens, such as Fusarium24. In summary, the application of HA has a positive effect on soil health and plant growth. However, the impact of appropriate HA concentration on the Sanqi growth and soil health remains unclear. Excessive addition of organic fertilizer can lead to the accumulation of heavy metals25, disruption of soil element proportions26, and the occurrence of root rot2, which can further reduce crop yield and quality. Hence, screening the appropriate concentration of HA is crucial for increasing the yield of Sanqi and mitigating the challenges of CCOs.

Organic fertilizers with varying concentrations can cause differences in plant growth, soil nutrients, and microbial communities2. The key objective was to examine the effects of externally supplemented HA at varying concentrations on the growth of Sanqi and the overall health of the soil. To achieve this, we analyzed the diversity and community structure of bacteria and fungi in the Sanqi soils using high-throughput sequencing technology. Additionally, we examined the growth of Sanqi, along with soil edaphic factors, to understand the relationship among Sanqi growth, soil edaphic factors, and the stability of the soil microbial network. Our study serves two purposes—firstly, to identify the most suitable HA concentration for Sanqi growth, and secondly, to elucidate the relationship among Sanqi growth, soil edaphic factors, and soil microbes. The findings can provide valuable theoretical support for applying HA to Sanqi, thus laying a foundation for enhancing Sanqi yield.

Results

Effect of different concentrations of HA on plant growth

Compared to the control group (CK), the fresh and dry weights of Sanqi were significantly increased at moderate and high concentrations. The most significant increase was observed at the moderate concentration, whereas the fresh and dry weights of Sanqi did not show any significant difference at low concentration (Fig. 1).

Figure 1
figure 1

Fresh weight and dry weight of Sanqi under HA conditions. (a), fresh weight, (b), dry weight.

Full size image

Effect of HA on soil edaphic factors

The TN content significantly decreased at low concentration compared to the control group, but there was no significant difference at moderate and high concentrations. On the other hand, the NO3-−N content significantly increased at moderate and high concentrations, while there was no significant difference at low concentration. Similarly, there was no significant difference in the content of TP and NH4+−N at different concentrations (Fig. 2).

Figure 2
figure 2

Soil edaphic factors of Sanqi under HA conditions. (a), TN; (b), TP; (c), NH4+−N; (d), NO3–N.

Full size image

Analysis of bacterial and fungal α-diversity

There was no significant difference in the Shannon index of bacteria at different concentrations. However, the Shannon index of fungi increased significantly at low concentrations, while there was no significant difference at moderate to high concentrations (Fig. 3).

Figure 3
figure 3

Changes in diversity index of bacteria and fungi under HA conditions. (a) bacteria (b) fungi.

Full size image

Community structure and β-diversity analysis of bacteria and fungi

The bacterial and fungal composition differences were explained by the PC1 and PC2 axes, which accounted for 28.44 and 17.96% (R = 0.59, P = 0.001), and 25.19 and 23.15% (R = 0.68, P = 0.001), respectively. Compared with CK, the bacterial and fungal community structures were most different at moderate rather than high or low concentrations, as shown in (Fig. 4a,b). Moreover, there was no significant difference in the β-diversity of bacteria at different concentrations compared to CK. The β-diversity of fungi was not significantly different at low and moderate concentrations, but decreased significantly at high concentrations (Fig. 4c,d).

Figure 4
figure 4

Principal co-ordinates analysis (PCoA) and β diversity analysis of soil bacteria and fungi. (a,c) bacteria; (b,d) fungi.

Full size image

Analysis of community composition at the bacterial and fungal phylum and genus levels

The bacterial and fungal communities were analyzed at the phylum and genus levels under HA treatment conditions. Proteobacteria (44.83%), Acidobacteria (16.89%), Actinobacteria (10.47%), and Chloroflexi (9.23%) were the top four bacterial phyla, accounting for 81.43% of all bacteria. A total of 30 phyla, 66 classes, 151 orders, 230 families, 388 genera, and 786 species were identified. Compared to CK, different concentrations of HA decreased the abundance of Proteobacteria and Acidobacteria, while increasing the abundance of Actinobacteria and Chloroflexi (Fig. 5a).

Figure 5
figure 5

The dominant phyla (a,b) and genus (c,d) of soil microbes under HA conditions. (a,c), bacterial phyla levels; (b,d), fungal phyla levels.

Full size image

The top four fungal phyla were Ascomycota (50.02%), Mortierellomycota (25.74%), Basidiomycota (21.96%), and unclassified_k__Fungi (1.88%), accounting for approximately 97.72% of all fungi. A total of 13 phyla, 30 classes, 66 orders, 125 families, 195 genera, and 283 species were identified. Different concentrations of HA increased the abundance of Ascomycota and Mortierellomycota, while decreasing the abundance of Basidiomycota (Fig. 5b).

The relative abundance of the top ten bacterial genera were: Bradyrhizobium (9.74%), unclassified_f__Xanthobacteraceae (9.4%), unclassified_o__Elsterales (4.36%), unclassified_o__Acidobacteriales (4.05%), Burkholderia-Caballeronia-Paraburkholderia (3.79%), unclassified_o__Subgroup_2 (3.31%), Candidatus_Udaeobacter (2.85%), unclassified_c__AD3 (2.77%), Candidatus_Solibacter (2.64%), unclassified_f__Acetobacteraceae (1.8%), accounting for approximately 44.71% of the total bacteria. The abundance of unclassified_f__Xanthobacteraceae, Burkholderia-Caballeronia-Paraburkholderia, unclassified_o__Elsterales, unclassified_o__Subgroup_2, Candidatus_Solibacter decreased at different concentrations, while the abundance of Bradyrhizobium, unclassified_c__AD3, unclassified_o__Acidobacteriales, Candidatus_Udaeobacter increased. The moderate concentration of HA significantly increased the abundance of Bradyrhizobium and Burkholderia-Caballeronia-Paraburkholderia, while significantly decreasing the abundance of unclassified_f__Xanthobacteraceae (Fig. 5c).

The most abundant fungal genera, in descending order, were Mortierella (25.71%), unclassified__f__Thelephoraceae (12.4%), Exophiala (7.39%), unclassified__p__Ascomycota (5.65%), Penicillium (5.33%), Cladophialophora (3.77%), Solicoccozyma (3.72%), Trichophaea (3.27%), Talaromyces (2.72%), Devriesia (2.33%), accounting for approximately 72.29% of the total fungal (genus) species. The concentrations of Mortierella, Exophiala, unclassified_p__Ascomycota, Penicillium, Solicoccozyma increased, while the abundance of unclassified_f__Thelephoraceae, Trichophaea decreased at different concentrations. Specifically, low concentration of HA significantly increased the abundance of Mortierella and significantly decreased the abundance of unclassified_f__Thelephoraceae. High concentration of HA significantly increased the abundance of unclassified_f__Thelephoraceae, unclassified_p__Ascomycota and Solicoccozyma (Fig. 5d).

Relationship between soil edaphic factors and microbial communities

The RDA analysis showed that in CK conditions, the bacterial α diversity had a significant negative correlation with the dry weight of Sanqi and NO3-–N. The fungal α diversity was negatively correlated with TN, TP, and NO3-–N, and positively correlated with the fresh weight and NH4+–N. TN, TP. NO3-–N has a significant impact on bacterial community composition. In contrast, soil edaphic factors have no effect on fungal community composition (Fig S1). In addition, under the condition of HA application, the bacterial α diversity was positively correlated with the dry and fresh weights of Sanqi, NO3-–N, and negatively correlated with NH4+–N. Conversely, the fungal α diversity negatively correlated with the dry and fresh weights of Sanqi, TN, TP, and NO3-–N (Fig. 6a,b). The bacterial community composition was significantly affected by NH4+–N (P < 0.05), whereas the fungal community composition was significantly affected by the fresh weight of Sanqi, TN, TP, NH4+–N, and NO3-–N (P < 0.05) (Fig. 6c).

Figure 6
figure 6

Relationships among α diveisity, soil edaphic factors and community composition under the condition of HA. Results from redundancy analysis of bacteria and fungi (a,b) and mantel analysis (c) are shown under different concentrations of organic fertilizer conditions. Red indicates a positive correlation, and blue indicates a negative correlation (*P < 0.05, **P < 0.01).

Full size image

Direct and indirect relationship among plants, soil microbial community and microbial network stability

The study evaluated the robustness of the soil microbial network by applying different concentrations of HA. The results showed that moderate concentrations of HA increased the stability of both bacterial and fungal networks compared to the control group. However, high concentrations of HA had the opposite effect and decreased the stability of the microbial networks. In addition, low concentrations of HA decreased the stability of bacterial networks while increasing the stability of fungal networks, as illustrated in (Fig. 7).

Figure 7
figure 7

Changes in microbial network stability under different concentrations of HA. The proportion of remaining nodes in the bacterial (a) and fungal (c) network after random removal some nodes. The proportion of remaining nodes in the bacterial (b) and fungal (d) network after random deletion of 50% of the nodes.

Full size image

The SEM analysis results revealed that the bacterial network stability was directly influenced by the fresh weight of Sanqi, TP, NO3-–N and bacterial community composition when subjected to moderate HA concentration conditions. Furthermore, TN, bacterial α diversity, and fresh weight of Sanqi can affect the bacterial community composition, further impacting the bacterial network stability. Notably, NO3-–N were the direct factors influencing the fungal network stability under CK treatment. In addition, the fresh weight of Sanqi and the fungal community can directly affect the stability of the fungal network. The standardization effect showed that the fresh weight of Sanqi and the bacterial and fungal community composition were the strongest predictors of the microbial network stability (Fig. 8; Tables S2–5).

Figure 8
figure 8

Effects of the fresh weight of sanqi, soil edaphic factors (TN, TP, NH4+−N, NO3-−N), microbial α diversity, and community composition on stability by structural equation model (SEM). (A) Red and green arrows represent both positive and negative correlations, respectively (*P < 0.05, ** P < 0.01, *** P < 0.001). (B) Bar graphs are the standardized effects from SEM on the stability. (a, b), bacterial stability under CK and moderate concentrations, respectively. (c, d), fungal stability under CK and moderate concentrations, respectively.

Full size image

Discussion

The concentration of organic fertilizer can significantly affect both the growth and quality of plants13. The appropriate concentrations of organic fertilizer can improve the soil environment, boost fertility, and facilitate plant growth. However, excessive or insufficient amounts of organic fertilizer can harm both the soil and plants. Our results indicated that moderate concentrations of HA, but not other concentrations (low and high), can significantly increase the dry and fresh weight of Sanqi, which was similar to previous research that moderate concentrations of HA can promote the growth of Vicia faba14 and Vigna radiat18. On the one hand, HA can increase the content of soil metabolites (carbohydrates), thereby increasing plant yield12,13, and HA can also increase enzyme activity, inhibit disease development27,28, and enhance plant stress tolerance16. On the other hand, although HA has been demonstrated to mediate and improve nutrient uptake by plants13, it also increases soil nutrient levels such as nitrogen29,30, alters soil microbial composition, and further regulates soil nutrient cycling31. Previous studies have shown that the application of HA increases the content of total nitrogen, total phosphorus, total potassium (TK), and soil organic matter (SOM), etc. in soils such as Arachis hypogaea and Gossypium hirsutum15,20. Differently, our results indicated that no significant differences were observed in the content of TP and NH4+–N among the different treatments. It should be noted that, compared with CK, the content of NO3-–N increases at moderate and high concentrations, and adequate levels of NO3-–N are beneficial for nutrient uptake by plants32. In addition, HA, being a slow-release organic fertilizer, can continue to provide nutrients to the soil11. Therefore, we speculated that HA promoted the increase of NO3-–N content and leading to an increase the dry and fresh weight of Sanqi.

Our research indicated that moderate concentrations of HA had the greatest impact on the community structure of soil bacteria and fungi, as opposed to other concentrations. This finding is inconsistent with the results of Li et al.15 and could be attributed to differences in plant species and concentrations of HA application33. HA can affect both the physical–chemical properties of soil and the microbial community structure, as it serves as a source of nutrients (such as reducing sugars, organic acids, and amino acids) for microbes34,35. Therefore, we hypothesized that varying concentrations of HA provide distinct nutrients for microbes, leading to changes in the microbial community structure. Moreover, soil bacterial and fungal α diversity and β diversity did not exhibit significant changes at moderate and high concentrations, unlike the results of Maji et al.17, which reported that HA increased bacterial and fungal diversity in Pisum sativum soil. Soil physical–chemical properties have been shown to influence changes in microbial diversity, according to Jiang et al.36. Our correlation analysis revealed that microbial diversity was significantly correlated with TN, TP and NH4+–N (Table S6), but the contents of TN, TP and NH4+–N did not change significantly at moderate and high concentrations. Therefore, we speculated that the lack of change in diversity may be due to the absence of change in the associated soil edaphic factors.

Different concentrations of HA decreased the abundance of Acidobacteria, which is consistent with the findings of Chen et al.37 on nitrogen fertilizer application. Applying organic fertilizer can increase soil pH38, which is negatively correlated with Acidobacteria52. Simultaneously, HA also reduced the abundance of Proteobacteria and Basidiomycota. Research has shown that Sanqi roots also secrete organic acids into the soil after the application of organic fertilizer39, which can reduce the abundance of Proteobacteria and Basidiomycota and further inhibit their ability to decompose organic matter and lignin8. Additionally, HA also increased the abundance of Actinobacteria, Chloroflexi, and Ascomycota in the soil. Research has demonstrated that an increase in elemental nitrogen can lead to a higher growth rate of Actinobacteria, Chloroflexi, and Ascomycota in soil33,40,41. This is due to the fact that HA increases the NO3-–N content of the soil, which in turn increases the abundance of Actinobacteria, Chloroflexi, and Ascomycota, further enhancing the rate of soil carbon and nitrogen cycling40, accelerating the mineralization of SOM42, and providing a carbon source for the soil33.

However, moderate concentrations of HA significantly increased the abundance of Bradyrhizobium and decreased the abundance of Xanthobacteraceae, which is similar to previous findings that organic fertilizers are effective in improving the community structure of microbes and increasing the abundance and number of beneficial flora43. Studies have shown that the abundance of Bradyrhizobium in healthy rhizosphere soil is considerably higher than that in diseased rhizosphere soil, and it has been found to have the ability to fix nitrogen44. Xanthobacteraceae is a plant pathogenic bacterium that can lead to leaf spot and wilt45. At the same time, the application of HA effectively promotes plant growth and increases the number of beneficial microbes in and around plant roots4. In summary, the moderate concentration of HA increased the abundance of beneficial microbes, decreased the abundance of harmful microbes, regulated nutrient cycling more effectively, and further influenced the structural changes of microbial communities, thereby promoting the growth of Sanqi. Notably, high concentrations of HA also increased the abundance of some fungal genera (unclassified_f__Thelephoraceae, unclassified_p__Ascomycota, Solicoccozyma). This may be due to excess HA stimulates microbial growth and increases the microbial community, leading to enhanced competition for nutrients4.

In agricultural ecosystems, the stability of microbial networks represents an increase in plant yield, an accelerated rate of organic matter decomposition, and a reduction in the number of pathogen46,47. Soil microbial communities can be selected based on the availability of different resources48,49. Our study results indicated that the stability of soil bacteria and fungi can vary based on different concentrations of HA, with an moderate concentration significantly increasing the stability of soil microbes. This finding aligns with previous studies that found the addition of organic nitrogen and phosphorus fertilizers can improve the network stability of soil bacteria and fungi in crops such as Oryza sativa and Triticum aestivum, respectively33,37. The specific concentrations of organic fertilizer improve the stability of microbial networks by influencing SOM and the network complexity of bacteria and fungi, thereby enhancing the stability of plant ecosystems37,50. Additionally, soil microflora plays a crucial role in driving energy flows and material cycles in ecosystems, which can alter soil fertility and crop yields51.

SEM analyses revealed that the fresh weight of Sanqi and the bacterial/fungal community composition were the direct factors influencing the network stability of bacteria and fungi. The roots of Sanqi secreted various substances such as saponins, flavonoids, and organic acids that can affect the structure of the microbial community38, ultimately affecting microbial stability. We hypothesized that the moderate concentration of HA increased the content and variety of metabolites that provide energy for microbes, further promoting the stability of microbial networks. Additionally, organic fertilizers can enhance community associations by significantly influencing microbial interactions52, thereby increasing microbial network stability53. In natural systems, the stability of microbial networks ten to increase as environmental stress decreases54. Therefore, we speculated that appropriate concentration (moderate) of HA have improved the quality of Sanqi, promoted the interrelationships of microbial communities, thereby enhancing the stability of microbial networks and further strengthening the stability of the Sanqi from the forest understorey ecosystem.

Conclusions

The findings indicated that the fresh and dry weight of Sanqi was significantly increased by moderate concentration (4 ml/L) of HA, but not by low/high concentrations. The content of TP and NH4+–N remained unchanged with different concentrations of HA, while the content of NO3-–N was significantly increased with moderate/high concentrations of HA. Moderate levels of HA application exhibited the significant impact on the composition of soil bacterial and fungal communities. Furthermore, different concentrations of HA were found to decrease the abundance of Proteobacteria, Acidobacteria, and Basidiomycota, while increasing the abundance of Actinobacteria, Chloroflexi, Ascomycota, and Mortierellomycot. Specifically, the moderate concentration of HA also increased the abundance of beneficial bacteria (Bradyrhizobium), decreased the abundance of harmful bacteria (Xanthobacteraceae) and in particular improved the stability of bacterial and fungal networks. Mantel analysis showed that under HA conditions, NH4+–N had a significant impact on bacterial community composition (P < 0.05), while fungal community composition was more influenced by the fresh weight of Sanqi, TN, TP, NH4+–N, and NO3-–N. SEM results showed that the fresh weight of Sanqi and the community composition of microbes were the most significant predictors of microbial network stability at moderate concentrations of HA. It can be concluded that the moderate concentration of HA is the optimal choice for promoting the growth of Sanqi. In the next step, we should focus on long-term fertilization to further understand the effects of humic acid on the quality and soil health of Sanqi.

Materials and methods

Study sites and sanqi planting

The study sites were located in Dongshan Town, Mile City, Honghe Hani and Yi Autonomous Prefecture, Yunnan Province (103°3′22″E, 23°49′32″N), China. The altitude of the study sites ranges from 2045 to 2315 m. The region has a subtropical monsoon climate with an average annual precipitation of 935.4 mm and an average annual temperature of 12.8 °C. The soil type in the region is typical red soil, and the pine species is Pinus armandii, with an average height of 5–10 m and an average diameter of 10 cm.

According to the method described in the research by Hei et al.10, the Sanqi seedlings, which are one year old, were transplanted into the pine forest of Pinus armandii in January 2022. The seedlings were planted 10 cm apart in rows spaced 15 cm apart with a planting depth of 3–5 cm. Once transplanted, the Sanqi seedlings were covered with soil (2–5 cm) and then covered with pine needles.

Experimental design

The primary raw material of humic acid (HA) is the natural substance extracted from Atlantic leaf alginate, which is rich in alginate and other plant growth promoting substances. Meanwhile, potassium fulvic acid, which is rich in functional groups and various mineral elements, is added to the formula. Yunnan Jiayinong Agricultural Technology Co., LTD (Shandong, China) supplies HA with the following specifications: HA ≥ 30 g/L; N + P2O5 ≥ 200 g/L; PH: 7; Water insoluble matter ≤ 50 g/L; S ≤ 30 g/L; Cl ≤ 30 g/L; Na ≤ 30 g/L.

In this experiment, a random design was used, and 24 plots of equal size (10 m × 10 m) were selected. Different concentrations of HA (2, 4, and 6 ml/L) were added to Sanqi plants in the treatment groups, while the control group received distilled water. The HA was diluted at a ratio of 1:1000. During the vigorous growth period of Sanqi (July), we applied the HA to the roots of the plants at 200 ml per plant, ensuring that the roots were thoroughly watered. After 30 and 60 days of HA application, we collected Sanqi plants and rhizosphere soil, respectively. A total of 24 plant and soil samples were collected-4 (concentration) × 2 (time) × 3 (replication). To preserve the samples, they were placed in dry ice and liquid nitrogen and transported to the laboratory immediately after sampling. The rhizosphere soil samples were collected using the shaking soil method7.

Fresh weight, dry weight, and soil edaphic factors of sanqi

The sanqi plant was thoroughly washed with water and cleaned using filter paper. Then, it was weighed using a balance and dried for 48 h in an oven at 70℃ before being weighed again to obtain the dry weight. To determine the total phosphorus (TP), total nitrogen (TN), nitrate nitrogen (NO3-−N), and ammonium nitrogen (NH4+−N) content, a continuous flow analyzer (SEAL Analytical AA3) was used.

DNA extraction and PCR analysis

Soil samples weighing 0.5 g were used to extract DNA using the Fast DNA® Spin Kit for Soil (MP Biomedicals, Santa Ana, CA, USA). The quality and concentration of DNA were assessed using a NanoDrop® ND-2000 spectrophotometer (Thermo Scientific Inc, Waltham, MA, USA) and 1.0% agarose gel electrophoresis. The bacterial 16S rRNA gene (V3-V4) and the fungal ITS gene were amplified with the primer pairs 338F/806R and ITS1F/ITS2R (Table S1), respectively. All samples underwent triplicate amplification. The PCR product was purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) after being extracted from a 2% agarose gel.

High-throughput sequencing

The amplified DNA fragments were purified, pooled, and then sequenced using the Illumina MiSeq PE300 platform (Illumina, San Diego, USA). The sequences were quality-filtered using fastp version 0.19.6, and then merged using FLASH version 1.2.78. The optimized sequences were then clustered into operational taxonomic units (OTUs) using UPARSE 7.1 at a sequence similarity level of 97%55. The raw read sequences obtained from 24 soil samples were 1 366, 954 for bacteria and 1 454, 752 for fungi. And they were clustered into 8515 and 2364 OTUs, respectively. The bacterial and fungal OTUs were assigned to the SILVA database and the UNITE database10.

Statistical analysis

The study examined the α diversity indices (Shannon) using Mothur v1.30.156. One-way ANOVA followed by Duncan's test was conducted using IBM SPSS statistical 23 software (SPSS, Inc., Chicago, IL, USA) to determine significant differences in soil edaphic factors and α diversity. A vegan v2.5-3 package was used to perform Principal coordinate analysis (PCoA) based on the Bray-Curtis distance to profile the differences in the treatments. Circos (https://cloud.oebiotech.cn/task/detail/rna_chord/, accessed on 22 March 2024) was utilized to analyze the composition of microbes. R software (version 3.3.1) was used to figure out the column plots of the microbial community with abundance at genera levels. Redundancy analysis (RDA) was used to determine the relative contributions of the correlation between soil characteristics and microbial community composition. Mantel analysis was performed to determine the effects of soil edaphic factors and α diversity on the microbial community composition. The remaining proportion of nodes and robustness were analyzed to assess network stability using R software (version 3.3.1). Networks with higher remaining node proportion and robustness were more stable57.

Structural equation models (SEM) were used to evaluate the direct and indirect relationships among plant growth, soil edaphic factors (TN, TP, NH4+−N, NO3-−N), microbial α diversity, community composition, and network stability. The effect of moderate HA concentration on the stability of the soil microbial network was determined by the path standardization coefficient and correlation P-value. The goodness of fit index (GFI) was used to assess the model’s fit with a higher GFI value, indicating a more reliable structural equation model58. Wu et al.59 concluded that the microbial (bacterial and fungal) networks stability is determined by the slope between the natural connectivity and the proportion of removed nodes.

Research involving plants statement

Experimental research and field studies on plants, including the collection of plant material, comply with relevant all institutional, national, and international guidelines and legislation for cultivated plants.

Ethical approvals

This study was approved by China Agriculture Research System for permission to collect plant specimens. The plant samples were formally identified as Sanqi (Panax notoginseng (Burk.) F.H.Chen) by Professor Shu Wang, College of Landscape and Horticulture, Southwest Forestry University. The Sanqi samples were stored in Key Laboratory of Underforest Resource Protection and Utilization in Yunnan Province, College of Landscape and Horticulture of Southwest Forestry University, Kunming. The Sanqi sample numbers are as follows: A_CK_1, A_CK_2, A_CK_3, A_L_1, A_L_2, A_L_3, A_M_1, A_M_2, A_M_3, A_H_1, A_H_2, A_H_3, S_CK_1, S_CK_2, S_CK_3, S_L_1, S_L_2, S_L_3, S_M_1, S_M_2, S_M_3, S_H_1, S_H_2, S_H_3.